1185 Predictors of Hospital Readmission Among Patients With Obscure Gastrointestinal Bleeding Following Inpatient Capsule Endoscopy

Abstract

INTRODUCTION: Patients with obscure gastrointestinal bleeding (OGIB) often require frequent hospital admissions and endoscopic procedures to identify and treat bleeding. Capsule endoscopy (CE) provides a novel approach to evaluate OGIB. The role of inpatient CE in hospital readmissions has not been previously reported. We sought to identify patient and clinical factors associated with hospital readmission among patients with OGIB following inpatient CE. METHODS: We conducted a retrospective review of adult patients undergoing inpatient CE at a tertiary referral, academic medical center between 4/1/17 and 4/1/19 to evaluate for predictors of readmission. We reviewed patient demographics, medical history, medications, hospital course, results of CE, and 30-day readmission rates. Arteriovenous malformation (AVM) risk factors were defined as having comorbid cardiac valvular disease or advanced chronic kidney disease (stage 3-5). A positive CE was defined as having abnormal small bowel pathology or bleeding. RESULTS: 86 patients were included. Overall, 64% were male with a median age of 69 years (Table 1). The main indication for inpatient CE was overt bleeding (59%). Complete CE was noted in 70 patients (81%), and positive CE in 48 patients (56%). Overall, 17 patients (19.8%) were readmitted within 30 days of index admission. Length of stay (LOS) on index admission was longer among those readmitted (Mean LOS 10 days vs 7.9 days, P = 0.07). Readmission was not related to age, gender, complete CE, positive CE or bleeding on CE. There was a trend for readmission among patients who had previously been admitted within 30 days (41% vs 22%, P = 0.1). There was a trend towards readmission for patients with diabetes mellitus (DM) and AVM risk factors. Medication use was not associated with readmission. CONCLUSION: Patient with OGIB have high rates of hospital admission. The use of inpatient CE can provide guidance on subsequent therapy in patients with OGIB, with the potential to decrease hospital admissions. We found high rates of 30-day hospital readmission among patients with OGIB after inpatient CE. No significant predictors of readmission were noted, though several factors, including increased LOS on index admission, DM, AVM risk factors, and recent admission trended towards readmission. Larger studies are necessary to fully elucidate predictors of readmission. To our knowledge, this is the first study to address the effect of inpatient CE and hospital readmissions.